beomi commited on
Commit
c90eadc
1 Parent(s): 55bc6ef

update with 10 epoch model

Browse files
all_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 1.0,
3
- "train_loss": 5.64251610059892,
4
- "train_runtime": 333.7038,
5
  "train_samples": 7931,
6
- "train_samples_per_second": 23.767,
7
- "train_steps_per_second": 2.973
8
  }
 
1
  {
2
+ "epoch": 10.0,
3
+ "train_loss": 5.131693365496973,
4
+ "train_runtime": 5547.9529,
5
  "train_samples": 7931,
6
+ "train_samples_per_second": 14.295,
7
+ "train_steps_per_second": 1.788
8
  }
config.json CHANGED
@@ -3,7 +3,7 @@
3
  "exBertForMaskedLM"
4
  ],
5
  "attention_probs_dropout_prob": 0.1,
6
- "base_model": "",
7
  "bos_token_id": 0,
8
  "eos_token_id": 2,
9
  "ex_hidden_size": 252,
 
3
  "exBertForMaskedLM"
4
  ],
5
  "attention_probs_dropout_prob": 0.1,
6
+ "base_model": "beomi/kcbert-base",
7
  "bos_token_id": 0,
8
  "eos_token_id": 2,
9
  "ex_hidden_size": 252,
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:22b79319a7b3d601effaa9452c0128d7e3ba5c7395cc26f885bd79257f03aa7d
3
  size 672770099
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94965753ac746cf9ffc407b17007d2fb3ab928d9ee345a5ec1bd959f1bdbb3a8
3
  size 672770099
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 1.0,
3
- "train_loss": 5.64251610059892,
4
- "train_runtime": 333.7038,
5
  "train_samples": 7931,
6
- "train_samples_per_second": 23.767,
7
- "train_steps_per_second": 2.973
8
  }
 
1
  {
2
+ "epoch": 10.0,
3
+ "train_loss": 5.131693365496973,
4
+ "train_runtime": 5547.9529,
5
  "train_samples": 7931,
6
+ "train_samples_per_second": 14.295,
7
+ "train_steps_per_second": 1.788
8
  }
trainer_state.json CHANGED
@@ -1,37 +1,145 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 1.0,
5
- "global_step": 992,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
  "epoch": 0.0,
12
- "learning_rate": 4.994959677419355e-05,
13
- "loss": 10.8889,
14
  "step": 1
15
  },
16
  {
17
  "epoch": 0.5,
18
- "learning_rate": 2.4798387096774196e-05,
19
- "loss": 5.9914,
20
  "step": 500
21
  },
22
  {
23
- "epoch": 1.0,
24
- "step": 992,
25
- "total_flos": 2400440637256200.0,
26
- "train_loss": 5.64251610059892,
27
- "train_runtime": 333.7038,
28
- "train_samples_per_second": 23.767,
29
- "train_steps_per_second": 2.973
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
30
  }
31
  ],
32
- "max_steps": 992,
33
- "num_train_epochs": 1,
34
- "total_flos": 2400440637256200.0,
35
  "trial_name": null,
36
  "trial_params": null
37
  }
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 10.0,
5
+ "global_step": 9920,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
  "epoch": 0.0,
12
+ "learning_rate": 4.9994959677419356e-05,
13
+ "loss": 10.8957,
14
  "step": 1
15
  },
16
  {
17
  "epoch": 0.5,
18
+ "learning_rate": 4.7479838709677423e-05,
19
+ "loss": 5.955,
20
  "step": 500
21
  },
22
  {
23
+ "epoch": 1.01,
24
+ "learning_rate": 4.495967741935484e-05,
25
+ "loss": 5.2306,
26
+ "step": 1000
27
+ },
28
+ {
29
+ "epoch": 1.51,
30
+ "learning_rate": 4.243951612903226e-05,
31
+ "loss": 5.1627,
32
+ "step": 1500
33
+ },
34
+ {
35
+ "epoch": 2.02,
36
+ "learning_rate": 3.991935483870968e-05,
37
+ "loss": 5.1156,
38
+ "step": 2000
39
+ },
40
+ {
41
+ "epoch": 2.52,
42
+ "learning_rate": 3.7399193548387094e-05,
43
+ "loss": 5.1212,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 3.02,
48
+ "learning_rate": 3.487903225806452e-05,
49
+ "loss": 5.1012,
50
+ "step": 3000
51
+ },
52
+ {
53
+ "epoch": 3.53,
54
+ "learning_rate": 3.2358870967741936e-05,
55
+ "loss": 5.0911,
56
+ "step": 3500
57
+ },
58
+ {
59
+ "epoch": 4.03,
60
+ "learning_rate": 2.9838709677419357e-05,
61
+ "loss": 5.0799,
62
+ "step": 4000
63
+ },
64
+ {
65
+ "epoch": 4.54,
66
+ "learning_rate": 2.7318548387096775e-05,
67
+ "loss": 5.0627,
68
+ "step": 4500
69
+ },
70
+ {
71
+ "epoch": 5.04,
72
+ "learning_rate": 2.4798387096774196e-05,
73
+ "loss": 5.0703,
74
+ "step": 5000
75
+ },
76
+ {
77
+ "epoch": 5.54,
78
+ "learning_rate": 2.2278225806451614e-05,
79
+ "loss": 5.0741,
80
+ "step": 5500
81
+ },
82
+ {
83
+ "epoch": 6.05,
84
+ "learning_rate": 1.975806451612903e-05,
85
+ "loss": 5.0708,
86
+ "step": 6000
87
+ },
88
+ {
89
+ "epoch": 6.55,
90
+ "learning_rate": 1.7237903225806452e-05,
91
+ "loss": 5.0604,
92
+ "step": 6500
93
+ },
94
+ {
95
+ "epoch": 7.06,
96
+ "learning_rate": 1.4717741935483872e-05,
97
+ "loss": 5.0575,
98
+ "step": 7000
99
+ },
100
+ {
101
+ "epoch": 7.56,
102
+ "learning_rate": 1.2197580645161291e-05,
103
+ "loss": 5.08,
104
+ "step": 7500
105
+ },
106
+ {
107
+ "epoch": 8.06,
108
+ "learning_rate": 9.67741935483871e-06,
109
+ "loss": 5.0736,
110
+ "step": 8000
111
+ },
112
+ {
113
+ "epoch": 8.57,
114
+ "learning_rate": 7.15725806451613e-06,
115
+ "loss": 5.0615,
116
+ "step": 8500
117
+ },
118
+ {
119
+ "epoch": 9.07,
120
+ "learning_rate": 4.637096774193548e-06,
121
+ "loss": 5.0382,
122
+ "step": 9000
123
+ },
124
+ {
125
+ "epoch": 9.58,
126
+ "learning_rate": 2.1169354838709676e-06,
127
+ "loss": 5.0565,
128
+ "step": 9500
129
+ },
130
+ {
131
+ "epoch": 10.0,
132
+ "step": 9920,
133
+ "total_flos": 2.4004406372562e+16,
134
+ "train_loss": 5.131693365496973,
135
+ "train_runtime": 5547.9529,
136
+ "train_samples_per_second": 14.295,
137
+ "train_steps_per_second": 1.788
138
  }
139
  ],
140
+ "max_steps": 9920,
141
+ "num_train_epochs": 10,
142
+ "total_flos": 2.4004406372562e+16,
143
  "trial_name": null,
144
  "trial_params": null
145
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c726d8f9c0eb16add1fb83e2d0607a69ad2a6405e5ee5d0744223baa48c49873
3
  size 2479
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6917ddad71848fd3497a5660a048092850c3e5c1b501a35877f9fe4ee8114b13
3
  size 2479